REM Sleep Imposes a Vascular Load in COPD Patients Independent of Sleep Apnea

Figures & data

Table 1. Anthropometric and clinical characteristics in patients with COPD (n = 27) and 2 non-COPD control groups (n = 115). Shown are means ± standard deviation.

Variable	COPD (n = 27)	CVD (n = 72)	HC (n = 43)	Between group analysis^*	Post hoc analysis COPD vs CVD	Post hoc analysis COPD vs HC
Females	33%	38%	30%	n.s.	n.s.	n.s.
Age (years)	64.7 ± 11	58.7 ± 13	49.3 ± 12	<0.001	0.03	<0.001
Body Mass Index (kg/m2)	31.2 ± 8	33.6 ± 6	27.6 ± 3	<0.001	n.s.	0.013
Office Systolic Blood Pressure (mmHg)	134.5 ± 18	137.6 ± 20	128.5 ± 16	0.04	n.s.	n.s.
Office Mean Arterial Pressure (mmHg)	98.9 ± 11	100.3 ± 13	97.5 ± 11	n.s.	n.s.	n.s.
Office Diastolic Blood Pressure (mmHg)	80.2 ± 10	81.7 ± 12	82.0 ± 9	n.s.	n.s.	n.s.
Pulse (beats/minute)	75.5 ± 16	71.5 ± 9	74.2 ± 11	n.s.	n.s.	n.s.
Pulse Pressure (mmHg)	56.2 ± 15	55.9 ± 16	46.5 ± 12	0.003	n.s.	0.012
Rate Pressure Product (bpm^*mmHg)	10259 ± 2863	9850 ± 1993	9505 ± 1701	n.s.	n.s.	n.s.
Current/previous smoker	58%	28%	0%	0.001	0.023	0.001
Mean Pack Years	19.9 ± 25	6.4 ± 13	0	<0.001	<0.001	<0.001
Total Cholesterol (mg/dl)	208.4 ± 26	203.6 ± 42	217.3 ± 44	0.008	n.s.	n.s.
High Density Lipoprotein (mg/dl)	51.4 ± 17	48.8 ± 11	48.4 ± 13	n.s.	n.s.	n.s.
	Comorbidities (percentage of patients)
Arterial Hypertension	74%	100%	0%	<0.001	0.001	<0.001
Pulmonary Hypertension	4%	2%	0%	n.s.	n.s.	n.s.
Chronic Heart Failure	19%	13%	0%	<0.023	n.s.	<0.003
Coronary Artery Disease	15%	17%	0%	0.019	n.s.	0.009
TIA/Stroke	4%	7%	0%	n.s.	n.s.	n.s.
Diabetes Mellitus	19%	29%	0%	<0.001	n.s.	0.03
Hyperlipidemia	20%	44%	16%	0.007	n.s.	n.s.
Medication (percentage of patients)
Betablocker	30%	43%	0%	<0.001	n.s.	0.001
Renin Angiotensin Blocker	41%	60%	0%	<0.001	n.s.	<0.001
Diuretics	34%	20%	0%	0.002	n.s.	<0.001
Calcium Channel Bockers	7%	36%	0%	<0.001	0.05	n.s.
Lipid lowering drugs	19%	31%	0%	0.001	n.s.	0.018
Antidiabetic drugs (including insulin)	11%	22%	0%	0.051	n.s.	0.025
	Sleep, sleep disordered breathing and nocturnal hypoxia
Total sleep time (h)	5.0 ± 1.3	5.2 ± 1.2	5.1 ± 1.2	n.s.	n.s.	n.s.
REM sleep %	12.3 ± 9	13.8 ± 7	13.9 ± 8	n.s.	n.s.	n.s.
Total arousal index (n/h)	12.2 ± 14	15.8 ± 18	18.2 ± 15	n.s.	n.s.	n.s.
Apnea Hypopnea Index (n/h)	14.1 ± 16	20.4 ± 19	13.8 ± 16	n.s.	n.s.	n.s.
Oxygen Desaturation Index 4% (n/h)	14.4 ± 18	18.9 ± 18	9.8 ± 15	0.003	n.s.	n.s.
Mean overnight
SaO2 (%)	93.1 ± 3	94.5 ± 3	95.9 ± 2	<0.001	<0.024	<0.001
Minimum overnight SaO2 (%)	80.9 ± 6	80.4 ± 6	85.7 ± 5	<0.001	n.s.	0.008
Time in overnight SaO2<90% (min)	74.9 ± 127	32.3 ± 70	8.0 ± 22	<0.001	0.002	<0.001
Time in REM sleep SaO2<90% (min)	7.3 ± 17	4.6 ± 8	1.1 ± 3	0.021	n.s.	0.008
Time in NREM SaO2<90% (min)	47.3 ± 84	22.8 ± 53	5.9 ± 19	0.002	0.046	0.002
Time in Wake SaO2<90% (min)	23.0 ± 41	4.8 ± 16	1.2 ± 2	<0.001	<0.001	<0.001
AHI≥15 n/h	29.6%	52.8%	34.9%	0.051	0.04	n.s.
	Pulmonary function data
FVC (l)	2.75 ± 0.67	3.51 ± 1.14	4.31 ± 0.95	<0.001	0.003	<0.001
FVC % predicted	78 ± 19	94 ± 14	97 ± 12	<0.001	<0.001	<0.001
FEV1 (l)	1.49 ± 0.45	2.79 ± 0.88	3.49 ± 0.79	<0.001	<0.001	<0.001
FEV1% predicted	51 ± 26	96.4 ± 16	95.6 ± 12	<0.001	<0.001	<0.001
FEV1/FVC	0.54 ± 0.12	0.83 ± 0.08	0.81 ± 0.07	<0.001	<0.001	<0.001

COPD = Chronic Obstructive Pulmonary Disease, CVD = Cardiovascular Disease, AHI = Apnea Hypopnea Index, FVC = Forced Vital Capacity, FEV₁ = Forced Expiratory Volume in the first second, HC = Healthy Controls, bpm = beats per minute, mg = milligram, n = number, h = hour, mmHg = millimeter Mercury

^* = Between group analysis by ANOVA-, Kruskal Wallis-, or Chi² -test

Figure 1. Study flow chart. From a pool of polysomnographic sleep studies two analyses have been performed in the current study: Analysis 1 targteted overnight PPT in 142 patients divided in 3 subgroups; analysis 2 targeted sleep stage based PPT analysis performed in 12 COPD patients and 12 matched controls.

Figure 2. Association between pulse propagation time (PPT in milliseconds) and time below 90% oxygen saturation (in minutes) for the three study groups: Healthy Controls group, Cardiovascular Disease group (CVD) and Chronic Obstructive Pulmonary Disease group (COPD). Spearmen rho correlation factor (R) and significance level are shown.

Table 2. Independent predictors of mean overnight Pulse Propagation Time (PPT) obtained in 3 different General Linear Models for (a) 27 COPD patients, (b) 72 CVD patients and (c) 43 healthy controls. Beta values indicate the numerical influence on PPT for each factor listed in the table.

	Beta value: Influences on PPT in ms (Mean and 95% CI)			Statistics
Covariate	COPD	CVD	Healthy Controls	COPD	CVD	Healthy Controls
Age	−1.1 (−1.9 – −0.3)	−1.6 (−2.1 – −1.0)	−2.0 (−2.7 – −1.0)	0.003	<0.001	<0.001
BMI	n.s.	1.9 (0.7–3.0)	n.s.	—	0.001	—
Systolic BP	n.s.	0.6 (0.2–1.0)	n.s.		0.004	—
Diastolic BP	−1.2 (−2.2–−0.2)	−1.8 (−2.5 – −1.1)	n.s.	<0.05	<0.001	—
Time SaO2<90% (10 min)	1.7 (1.1–2.3)	n.s.	4.7 (0.9–8.5)	<0.001	—	0.015
Intercept	303 (188–417)	261 (190–331)	316 (212–420)	<0.001	<0.001	<0.001

Non-significant factors tested in the model: Arterial hypertension, diabetes mellitus, total cholesterol, and Oxygen Desaturation Index 4% or Apnea Hypopnea Index

NA = not applicable, n.s. = not significant.

Table 3. Anthropometric and clinical data in patients with COPD and matched controls. Shown are means ± standard deviation.

Parameter	Matched COPD patients (n = 12)	Matched controls (n = 12)	Statistics
Anthropometric and clinical data
Females	33%	17%	n.s.
Age (years)	62.6 ± 8	62.3 ± 11	Matching variable
Body Mass Index (kg/m^2)	32.8 ± 7	33.1 ± 6	Matching variable
Office Systolic Blood Pressure (mmHg)	138.3 ± 24	144.7 ± 20	n.s.
Office Mean Arterial Pressure (mmHg)	102.1 ± 16	104.4 ± 14	n.s.
Office Diastolic Blood Pressure (mmHg)	83.9 ± 14	84.2 ± 13	n.s.
Pulse (beats/minute)	79.7 ± 14	75.3 ± 18	n.s.
Pulse Pressure (mmHg)	54.4 ± 17	60.6 ± 14	n.s.
Heart Pressure Product (bpm*mmHg)	10978 ± 2943	11135 ± 4203	n.s.
Current or previous smoker	75%	33%	n.s.
Mean Pack Years	30.4 ± 28	10.8 ± 17	0.056
Total Cholesterol (mg/dl)	207.3 ± 30	196.5 ± 21	n.s.
High Density Lipoprotein (mg/dl)	48.3 ± 11	53.8 ± 17	n.s.
Comorbidities (percentage of patients)
Arterial Hypertension	75%	92%	n.s.
Pulmonary Hypertension	0%	0%	n.s.
Chronic Heart Failure	42%	0%	0.02
Coronary Artery Disease	17%	25%	n.s.
TIA/Stroke	8%	8%	n.s.
Diabetes Mellitus	25%	42%	n.s.
Hyperlipidemia	42%	33%	n.s.
Medication (percentage of patients)
Betablocker	33%	50%	n.s.
Renin Angiotensin Blocker	33%	66%	n.s.
Diuretics	42%	17%	n.s.
Calcium Channel Bockers	17%	42%	n.s.
Lipid lowering drugs	25%	42%	n.s.
Antidiabetic drugs (including insulin)	17%	33%	n.s.
Sleep related hypoxia and breathing disorders
Apnea Hypopnea Index (n/h)	20.1 ± 20	23.0 ± 21	Matching variable
Oxygen Desaturation Index 4% (n/h)	20.4 ± 22	19.4 ± 21	n.s.
NREM Oxygen Desaturation Index 2% (n/h)	25.1 ± 24	27.1 ± 24	n.s.
REM Oxygen Desaturation Index 2% (n/h)	36.6 ± 35	29.8 ± 17	n.s.
Mean overnight SaO2(%)	90.7 ± 3	94.5 ± 3	0.002
Minimum overnight SaO2 (%)	78.4 ± 4	80.4 ± 9	n.s.
Time in overnight SaO2<90% (min)	150.3 ± 157	28.6 ± 41	0.023
Time in REM SaO2<90% (min)	11.6 ± 23	4.8 ± 5	n.s.
Time in NREM SaO2<90% (min)	94.0 ± 104	20.4 ± 34	0.036
Time in Wake SaO2<90% (min)	44.5.6 ± 53	3.3 ± 5	0.021

Figure 3. Percentage of change in pulse propagation time (PPT, mean and standard error of the means (SEM)) from wakefulness (= 100%) in patients with COPD and matched non-COPD controls. Asterix indicate significant differences (all p < 0.05): Within group increase in PPT from wake to N3 (non-COPD controls) and PPT decrease from wake to REM (COPD patients); between group difference in REM sleep PPT between COPD patients and non-COPD controls. N1, N2, N3 and REM indicate the sleep stages NREM1-3 and REM. Please note, a decrease in PPT corresponds to an increase in arterial stiffness and vice versa.

Table 4. Correlations between the change in pulse wave propagation time from wakefulness to REM-sleep and daytime hemodynamic parameters.

	PPT change in % from wakefulness to REM sleep
Variable	COPD (n = 10)	Controls (n = 12)
Office SBP	−.920^*	.297
Office MBP	−.764^*	.179
Office DBP	−.326	.059
Office HR	−.453	−.301
Pulse Pressure	−.835^*	−.367
Rate Pressure Product	−.876^*	−.358

^* = significance value P is ≤ 0.01

COPD = Chronic Obstructive Pulmonary Disease (n = 10), Controls = age, BMI, and AHI matched patients without COPD (n = 12), SBP = Systolic Blood Pressure, MBP = Mean Blood Pressure, DBP = Diastolic Blood Pressure, HR = Resting Heart Rate, Pulse Pressure = SBP-DBP, Rate Pressure Product = HR^*SBP

Figure 4. Association between the change in arterial stiffness (PPT) from wakefulness to REM sleep and daytime systolic blood pressure at rest in COPD patients and matched controls.